Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Document Access
This document may be downloaded from www.EQ-Assess.org.nz in parts:
1 Part A – Assessment Objectives and Principles
2 Part B – Initial Seismic Assessment
3 Part C – Detailed Seismic Assessment
Dunning Thornton
Task Group Leaders Alastair Cattanach
Consultants
Stuart Oliver Holmes Consulting Group Lou Robinson Hadley & Robinson
Project Management was provided by Deane McNulty, and editorial support provided by
Ann Cunninghame and Sandy Cole.
Oversight to the development of these Guidelines was provided by a Project Steering Group
comprising:
Dave Brunsdon
Kestrel Group John Hare SESOC
(Chair)
Quincy Ma,
Gavin Alexander NZ Geotechnical Society NZSEE
Peter Smith
Funding for the development of these Guidelines was provided by the Ministry of Business,
Innovation and Employment and the Earthquake Commission.
Part C – Detailed Seismic Assessment
Contents
Contents i
DATE: JULY 2017 VERSION: 1
Part C – Detailed Seismic Assessment
Contents ii
DATE: JULY 2017 VERSION: 1
Part C – Detailed Seismic Assessment
C6.1 General
C6.1.1 Scope and outline of this section
This section provides guidance on the Detailed Seismic Assessment (DSA) of existing steel
framed buildings. It does not address earthquake damaged steel framed buildings or the
retrofitting of existing buildings.
ASCE 41-13 (2014). Seismic evaluation of existing buildings, American Society of Civil Engineers, and
Structural Engineering Institute, Reston, Virginia, USA.
Clifton, G.C. and Cowie, K. (2013). Seismic design of eccentrically braced frames, HERA Publication
P4001:2013.
Clifton, G.C. and Ferguson, W.G. (2015). Determination of the post-earthquake capacity of an eccentrically
braced frame seismic resisting system, The University of Auckland, report to the Natural Hazards Research
Platform.
Feeney, M.J. and Clifton, G.C. (2001). Seismic design procedures for steel structures, HERA Report R4-76,
Manukau City, NZ. HERA, 1995. To be read with Clifton, G.C.; Tips on Seismic Design of Steel Structures,
Notes from Presentations to Structural Groups mid-2000; HERA, Manukau City, 2000.
FEMA 273 (1997). NEHRP guidelines for the seismic rehabilitation of buildings, Federal Emergency
Management Agency, FEMA Report 273, Washington, DC.
FEMA 356 (2000). Prestandard and commentary for the seismic rehabilitation of buildings, Federal Emergency
Management Agency, FEMA Report 356, Washington, DC.
NZS 1170.5:2004. Structural design actions, Part 5: Earthquake actions - New Zealand, NZS 1170.5:2004.
Standards New Zealand, Wellington, NZ.
NZS 3404 Part 1:1997. Steel structures standard, incorporating Amendments 1 and 2, NZS 3404:1997.
Standards New Zealand, Wellington, NZ.
Concentrically braced frame A braced frame where the members are subjected primarily to axial forces
(CBF)
Connector An item within a connection that transfers forces from one member or
connection component to another (e.g. bolts, rivets and welds)
Detailed Seismic A quantitative seismic assessment carried out in accordance with Part C of
Assessment (DSA) these guidelines
Eccentrically braced frame A braced frame in which at least one end of each brace frames only into a
(EBF) beam in such a way that at least one stable, deformable link beam is
formed in each beam if the elastic limit of the frame is exceeded. In this
event, energy is dissipated through shear and/or flexural yielding in the link
beams (termed the active link regions) and the bracing members and
columns have sufficient capacity to remain essentially elastic.
Full restraint against lateral Restraint that effectively prevents lateral deflection and twist of a member
buckling (FLR)
Lateral force-resisting The part of a structural system that provides resistance to earthquake
system induced forces
Lateral restraint An element that prevents lateral movement of the critical flange of a
member
Local buckling A local instability involving a change of shape of the member cross section
along a relatively short length of member under compression
Moment resisting frame A building frame system in which lateral loads are resisted by shear and
(MRF) flexure in members and joints of the frame
Overstrength The maximum strength that a member or a connection can develop due to
variations in material strengths, and strength gain due to strain hardening,
if applicable
Plate slenderness The ratio of the critical unsupported width of a steel plate to the average
plate thickness
Probable capacity The expected or estimated mean capacity (strength and deformation) of a
member, an element, a structure as a whole, or foundation soils. For
structural aspects this is determined using probable material strengths. For
geotechnical issues the probable resistance is typically taken as the
ultimate geotechnical resistance/strength that would be assumed for
design.
Segment The length between adjacent cross sections which are fully, partially or
laterally restrained, or the length between an unrestrained end and the
adjacent cross section which is fully or partially restrained
Ultimate limit state (ULS) A limit state defined in the New Zealand loadings standard NZS 1170.5:2004
for the design of new buildings
XXX%ULS shaking Percentage of the ULS shaking demand (loading or displacement) defined
(demand) for the ULS design of a new building and/or its members/elements for the
same site.
For general assessments 100%ULS shaking demand for the structure is
defined in the version of NZS 1170.5 (version current at the time of the
assessment) and for the foundation soils in NZGS/MBIE Module 1 of the
Geotechnical Earthquake Engineering Practice series dated March 2016.
For engineering assessments undertaken in accordance with the EPB
methodology, 100%ULS shaking demand for the structure is defined in
NZS 1170.5:2004 and for the foundation soils in NZGS/MBIE Module 1 of
the Geotechnical Earthquake Engineering Practice series dated March
2016 (with appropriate adjustments to reflect the required use of NZS
1170.5:2004). Refer also to Section C3.
𝑎𝑎 Distance between the centre of connectors and a flange cleat angle leg
𝑎𝑎1 Distance between the centre of connectors and the top edge of a flange
cleat angle
𝑏𝑏1 Width of contact between beam flange and welds and column
𝐶𝐶s Factor that accounts for the potential for deterioration in performance of
CBFs with increasing inelastic demand
𝐸𝐸 Modulus of elasticity
Symbol Meaning
𝐺𝐺 Permanent action
ℎ Storey height
𝑙𝑙 Member length
𝑚𝑚 Number of braces
Symbol Meaning
𝑛𝑛 Number of connectors
𝑛𝑛 Number of storeys
𝑛𝑛1 Length obtained by a 45° dispersion though half of the depth of a column
𝑛𝑛2 Length obtained by a 1:2.5 dispersion though column flange and root
radius
Symbol Meaning
𝑄𝑄 Imposed action
𝑡𝑡 Thickness
𝑉𝑉bi Storey i beam seismic shear demand determined from beam probable
capacity
Symbol Meaning
∆ Displacement
𝜃𝜃 Chord rotation
𝜃𝜃i Angle between a brace and beam at the top end of the brace
Each of these factors is discussed below. Also refer to Appendix C6A for general guidance
on the typical pre-1976 steel building systems used in New Zealand.
Note:
The weld materials used and fabrication processes adopted were some of the minor factors
that led to brittle fractures of welded connections in over 200 buildings during the 1994
earthquake in Northridge, California.
Inadequate load paths through connections was also considered to be the principal cause of
most local failures in multi-storey steel buildings in the 2010/2011 Canterbury earthquake
sequence.
Note:
When undertaking a seismic assessment of a steel framed building, assessing load paths
through connections is likely to be the most important aspect of the evaluation process.
Column bases and hold down bolts are the elements most prone to severe localised loss
of material due to long term corrosion. There were several reported failures of
industrial structural systems in the 1987 Edgecumbe earthquake due to column failures at
the base from corrosion. In addition, reduction in member strengths due to corrosion was
reported as one of the main factors contributing to failure of braces during this earthquake
(Butcher et al., 1998).
Note:
A condition assessment, particularly of pre-1976 steel framed buildings, is recommended
as part of the DSA. Refer to Section C6.4 for more details.
Local buckling of steel members occurs due to plate slenderness, while lateral torsional
buckling of steel members occurs when there is inadequate lateral bracing of compression
flanges. The elastic resistance to lateral buckling of a steel member is influenced by several
factors such as: unbraced length of the compression flange, geometric and material
properties of the member, and moment gradient along the member.
Experimental evidences indicate that local plate buckling generally results in a gradual
degradation of strength and stiffness in compact cross sections, while lateral torsional
buckling causes a rapid loss of strength and stiffness (Gupta and Krawinkler, 1999). Local
buckling of slender members causes a rapid loss of section and hence member capacity.
Note:
When large ductility demands that may result in significant deterioration in member
strength and stiffness are likely, P-delta effects will be worsened.
Slab participation may induce column flexural yielding, column shear failure or beam shear
failure modes in steel MRFs, depending on the relative strength of the members and the
connections. Slab participation may also cause damage to floor slabs and compromise the
capacity of the floor system to transfer seismic demands to the lateral force resisting
members; although the evidence from the 2010/2011 Canterbury earthquake sequence is that
the influence on composite slabs (concrete on steel deck on steel or concrete supporting
beams) is minimal.
Note:
When the connections of a steel framed building are semi-rigid, slab participation may
considerably increase the stiffness and strength of the connections (Roeder et al., 1994).
Slab participation may be beneficial in such buildings if it does not result in localised
column failures.
Cast iron columns are found in some of the oldest New Zealand buildings and, until the early
1900s, were often used as gravity carrying elements. Cast iron is a low strength and brittle
material not suitable for use in a seismic-resisting system or in a gravity system that is
required to sustain significant deformations. The tensile strength of cast iron is significantly
less than its compressive strength due to the presence of voids and cracks within the iron
matrix (Rondal and Rasmussen, 2003). The consequence of these non-ductile characteristics
is that the performance of cast iron columns is likely to be poor if they are part of the lateral
force resisting system and/or are subjected to significant lateral displacements.
Cast iron columns can be dependably retained in an existing building if they are used as a
propped gravity column, with the supports for the beams assessed and reinforced if necessary
(e.g. with steel bands) to avoid local fracture under seismic-induced rotations. However, the
strength of a cast iron column cannot be determined using the provisions for steel columns
in these guidelines as cast iron has a different stress-strain relationship to steel. Guidance on
the assessment of cast iron columns can be found in Bussell (1997) and Rondal and
Rasmussen (2003).
Wrought iron was also used to a limited extent for structural members in early New Zealand
buildings. However, its use largely ended around the 1880s and 1890s as these items were
costly to manufacture. The principal disadvantage of wrought iron as a building material was
the small quantities made in each production item (bloom), being only 20-50 kg. This meant
that the use of wrought iron in structural members required many elements to be joined by
rivets.
Wrought iron has good compressive and tensile strength, good ductility, and good corrosion
resistance. The performance of wrought iron members is considered comparable to that of
steel members from the same era.
C6.2.8.2 Design
Despite their apparent advantage over other building types of the same era such as
unreinforced masonry buildings, steel buildings designed before the introduction of
NZS 4203:1976 suffer from the fundamental drawback of being not designed according to
capacity design procedures.
Note:
Pre-1976 design methods generally assumed an elastic response, with no consideration
given to likely failure modes and with no ductile detailing requirements to ensure that
potential plastic hinge regions can dependably accommodate earthquake induced ductility
demands. In addition, no attention was generally given to load paths through connections
under inelastic response. Structural members of these buildings that should remain elastic
to avoid undesirable failure mechanisms may not have the capacity to resist overstrength
actions originating from potential plastic hinge regions and slab participation.
Additionally, structural members and connections that are provided to resist gravity
induced loads may not have the capacity to accommodate earthquake induced
displacement demands; although most early gravity systems with bolted or riveted
connections are considered to have high ductility capacity but very limited strength.
The pattern of damage observed during the 1995 earthquake in Kobe, Japan indicates that
three factors play a significant role in ensuring a good overall seismic performance of a steel
frame building not designed following the capacity design method:
• The beam-column connections of the frames of a building should be able to retain their
shear and axial force carrying capacity when the connections are sustaining flexural
actions from earthquake demands.
• The inelastic demand in the columns should be kept to a minimum. This demand is
principally due to local buckling or crippling failure, and also to general plastic hinge
formation.
• The inelastic response of the building should be essentially symmetric in nature and not
lead to a progressive movement of the building in one direction only.
Note:
Details of the damage sustained during the Kobe earthquake are provided in
reconnaissance reports such as that by Park et al. (1995).
In buildings constructed before the 1950s the structural members of steel frames are usually
encased in lightly reinforced concrete as fire protection (refer to Figure C6.1). The
reinforcement of the encasement is often inadequate and poorly detailed (Bruneau and
Bisson, 2000), which results in a significant increase in stiffness and a relatively modest
increase in strength of the encased members.
Spalling of the encasement concrete, particularly in the end regions of members, has the
potential to increase the nonlinear demands in the steel members if they are required to be
loaded beyond yield.
Even older steel framed buildings constructed before the 1936 New Zealand standard model
building by-law introduced seismic design requirements typically contain beams that are
deeper than the columns. The frames of these buildings generally contain simple and semi-
rigid riveted connections that have a modest flexural capacity. In addition, these connections
generally exhibit poor energy dissipation capability with lack of adequate strength and
stiffness and may serve as the weakest link during inelastic earthquake demands. However,
the seismic performance of similar structures dating back to the 1906 San Francisco
earthquake has generally been high.
Consequently, steel framed buildings have been generally regarded as ductile and resilient
against earthquake induced collapse. However, the significant damage observed during the
Northridge (1994) and Kobe, Japan (1995) earthquakes emphasises the vulnerability of even
recently constructed steel framed buildings and the need for attention to load paths.
Although hundreds of MRF buildings suffered this unexpected overload form of connection
damage, most of the buildings displayed no visible signs of distress after the earthquake
(such as permanent lateral deflections); nor was there significant damage to non-structural
components and contents. However, the capacity of these buildings to resist further
earthquake induced demands was significantly compromised and costly repairs were
required.
The main reason for the unexpectedly poor performance was the inability of the load paths
between the beams and the columns of the frames to transfer actions generated by plastically
responding beams into the columns. The inadequacy of these load paths caused fractures of
the beam flange to column flange connections. The majority of the fractures were observed
to occur at the bottom beam-column flange connections due to slab participation. In some
instances these bottom fractures were even observed to trigger web connection failures
(Krawinkler, 1995). Refer to Figure C6.2.
Note:
Details of the damage sustained during the Northridge earthquake have been widely
reported in reconnaissance reports such as that by Norton et al. (1994).
Figure C6.2: Welded connection fracture modes observed during the 1994 Northridge
earthquake (Krawinkler, 1995)
The inadequacy of the load paths of “Pre-Northridge” connections meant that even the best
fabricated beam to column connections were not able to develop plastic hinges in beams that
exceeded a depth of approximately 360 mm. However, the following factors were
considered to have minor contributions to the failures of “Pre-Northridge” connections
(FEMA 355E, 2000):
• The welding practice was such that bottom flange weld passes were interrupted at beam
webs, resulting in weld defects that served as crack initiators.
• The configuration of the connections made detection of hidden bottom weld defects
difficult, particularly at the beam webs.
• The filler metal employed was typically developed for high deposition rate welding and
had very low notch toughness as a result.
• There was use of large size beams in buildings that had few lateral force resisting frames.
The deeper the beam, the greater the web contribution to flexural strength and therefore
the greater the likelihood of ductile overload of the beam flange to column flange
connection. The use of large size beams also meant higher deposition rate large welds
which were more prone to fractures than small size welds (Krawinkler, 1995).
• The mean yield strength of members fabricated in the 1980s was observed to be generally
significantly greater than the nominal values.
• The geometry of weld access holes was, in some cases, observed to hinder ease of filler
metal deposition and weld inspections.
Immediately after the Northridge Earthquake, the New Zealand Heavy Engineering
Research Association (HERA) and the University of Auckland looked at the possibility of
similar types of failures in New Zealand buildings and found no examples of this type of
construction. A series of large scale beam/column inelastic cyclic tests were performed on
typical New Zealand type MRF connections which showed that they were not vulnerable to
this type of failure (Butterworth, 1995).
Provided the beams adjacent to the panel zone did not exhibit any signs of yielding, the
yielding of the panel zone was not expected. The yielding of this panel zone was considered
to result from the combination of elevated levels of compression force in the columns due to
high vertical ground accelerations and the expected and significant bending demands
imposed on the adjoining beams.
The good performance of multi-storey EBF buildings in the Canterbury earthquake sequence
can be attributed to:
• the significant effects of soil-foundation-structure-interaction (on reducing the seismic
demand on the superstructure of these relatively heavy multi-storey buildings built on
soft soil (Storie et al., 2014))
• factors contributing to overstrength in steel frames such as actual yield strengths
significantly exceeding nominal values, modelling assumptions, etc.
• the contribution of the composite floor slab action to the shear resistance that was not
allowed for in the design of the frames, and
• the contribution of solid partition walls and non-structural items.
A fractured active link of the 12 storey EBF building is presented in Figure C6.4(a). This
active link appeared to have undergone at least one full cycle of web panel yielding prior to
fracture. The fracture appeared to have propagated from one top corner across the active link
region and resulted in significant residual deformations. Detailed evaluations of this and
other links in the EBF braced bay concerned showed that the Charpy impact energy of this
steel was well below that specified by NZS 3404:1997, with the material having a transition
temperature of around 12oC. This particular link also had a shear stud welded to the flange
immediately above the left hand visible stiffener, which is believed to have acted as a crack
initiation site.
(a) A fractured active link in a 12 storey (b) A fractured active link in a low-rise
building parking building
Figure C6.4: Fractured EBF active links during the February 22, 2011 Christchurch
earthquake (Clifton and El Sarraf, 2011)
Fractures of two active links in a low-rise EBF building (refer to Figure C6.4(b)) were
attributed to detailing/fabrication errors. The flanges of the two braces were observed to be
offset from the stiffeners of the active links. The offset lead to fracture of unstiffened
collector beam flanges located between the active link stiffeners and the flanges of braces.
In New Zealand, non-capacity designed (pre-1976) CBFs are typically X-braced, while very
few are believed to be V-braced. Pre-1976 CBFs in New Zealand were typically designed to
resist lower levels of lateral forces than required by NZS 1170.5:2004. In Kobe, several such
CBF buildings were reported to sustain buckled braces or failed connections during the 1995
earthquake. However, none of these buildings were reported to have collapsed (Clifton,
1996).
Most CBF buildings performed as expected during the 1994 Northridge earthquake, but with
no collapses reported. Similar to the connection weld fractures of MRFs discussed in
Section C6.3.2, fractures of brace-collector beam and column-base plate welded connections
were prevalent. In addition, excessive local buckling of thin-walled tubular braces of CBFs
was observed (Krawinkler, 1995).
Figure C6.5: Damaged CBFs in a single-storey car park building during the February 22,
2011 Christchurch earthquake (Clifton and El Sarraf, 2011)
Significant damage to a single-storey CBF building was observed during the Canterbury
earthquake sequence (refer to Figure C6.5). However, the connections of the CBFs to
the columns appeared to have been poorly detailed. One of the CBFs appeared to have
been connected to a column that had a non-ductile reinforced concrete extension (refer to
Figure C6.5(a)), while the welded connection of the second CBF did not appear to have been
designed by following capacity design principles (refer to Figure C6.5(b)).
Many of the portal frame buildings in Christchurch were industrial facilities designed to
resist high wind induced forces, which were typically the controlling design case. These
buildings typically have light roofs that are braced using light rod braces with proprietary
end fittings. A few fractures and thread stripping of the proprietary brace connectors were
reported following the February 22, 2011 earthquake (refer to Figure C6.6).
Figure C6.6: Proprietary brace connectors that failed during the February 22, 2011
Christchurch earthquake (Clifton and El Sarraf, 2011)
Figure C6.7: Roof bracing that failed during the February 22, 2011 Christchurch earthquake
(Clifton and El Sarraf, 2011)
In one building, failure of a roof bracing was observed following the February 22, 2011
earthquake (refer to Figure C6.7). This failure was considered to be a result of excessive
movements of tilt up panels that were likely to have been caused by ground liquefaction
(Clifton and El Sarraf, 2011).
Mechanical properties of the steelwork within existing structural steel framed buildings may
be determined from:
• drawings, specifications or other construction records
• historical steel grades and nominal strengths, and/or
• steel material tests.
The mechanical properties of structural steelwork are best determined from original
construction records supplemented by laboratory or in-situ tests of selected critical
components to confirm the expected steel grade.
If the source of steelwork can be confirmed from the designations on original construction
records, but the steel grade is not identified and testing is not practicable, default mechanical
properties corresponding to the source and age of the steelwork can be adopted from those
outlined in historical specifications. Refer to Appendix C6B for typical sources of historical
New Zealand structural steelwork.
In the absence of construction records, the source of a structural steelwork can be identified
from the mill markings generally present on historical structural steel sections and
from section geometric properties contained in literature on historical structural steelwork
(e.g. Bates, 1991; Bussell, 1997; and Ferris, 1954).
Note:
Older steelwork exhibits greater variability than modern steelwork. Accordingly, a
minimum degree of non-destructive testing is recommended to gain assurance of the
mechanical properties for the members in the primary structure. This is particularly the
case when the steel is “of unknown origin”.
If the steelwork cannot be identified from construction records, mill markings or section
geometric properties, the default yield strengths for steel “of unknown origin” provided in
Appendix C6B may be adopted.
If tensile tests are undertaken, default strengths corresponding to the grade, potential source
and age of the steelwork should be adopted from Appendix C6B.
Members with steel of unknown origin may exhibit non-ductile behaviour if all of the
following conditions apply:
• from an assessment of the strength hierarchy of the building, the members with steel of
unknown origin are the weakest links, not the connections, and
• the members with steel of unknown origin are located in an external steelwork or on the
cold side of the building envelope so that the members could be below their transition
temperature at the time of an earthquake, and
• a notch, a significant crack, or any stress raiser is present in a critical location.
If all these conditions contributing to potential member brittle responses are present and
potential brittle failure has not already been ruled out through physical testing, fracture
toughness tests should be undertaken on selected critical members as per Section C6.4.5 to
rule out potential brittle responses.
Note:
Another key concern with members with “steel of unknown origin” is the undefined upper
bound on yield strength, which may be significantly greater than the characteristic values.
Primary members of unknown origin may develop strengths that are significantly higher
than allowed for using overstrength factors. Large member overstrengths may lead to
overloading other aspects of the structure and loss of assumed hierarchical behaviours
and/or protection.
The use of cast iron from the 1880s until its discontinuance around 1910 was limited to
columns. Cast iron columns would have been used typically for gravity load carrying
purposes. These columns are typically “chunky” with thick sections, often having ornate or
complex profiles (fluted, plain hollow circular, or cruciform shaped). The surface of these
columns is typically pitted with small blowholes.
Wrought iron
If a building is constructed before 1900 and contains members built up from many short-
length I-sections, channels and/or flats, then the possible use of wrought iron in these
members should be considered. Guidance for the assessment of wrought iron members is
provided in Bussell (1997).
Note:
Detailed visual assessment criteria for iron and steel members are presented in Bussell
(1997).
Table C6.1: Probable strengths of historical cast iron and wrought iron
Material Tensile strength Compressive strength Modulus of elasticity
(MPa) (MPa) (GPa)
Note:
Cast iron and wrought iron are generally only found in buildings constructed prior to 1900.
Due to the lack of available specific data, the probable strength of cast iron and wrought
iron is taken as one half of published breaking strengths such as those by Fidler (1879).
Table C6.1 is based on the lower characteristic strength values published in 1879
(Bates, 1991).
From the 1960s on most rolled sections have been manufactured in Australia, while plates
and welded sections have been mainly produced in New Zealand.
The structural steel properties outlined in relevant historical standard specifications are
summarised in Appendix C6B. Default characteristic/nominal strengths are also provided
for steel of unknown origin.
Note:
The first New Zealand structural steel standard specifications are believed to be NZS 309
and NZS 310, published in 1941. These standards and their subsequent editions were
based on their British equivalents until the first joint AS/NZS standard specifications were
introduced in the mid-1990s. The joint specifications were revisions of previous
Australian standard specifications.
Table C6.2: Factors to convert lower characteristic material strengths to probable strengths
(based on tests undertaken by Baker, 1969; Erasmus, 1984; and Erasmus and Smaill, 1990)
Period Steel grade Factor
Tests should at least identify the likely steel grades. They should also identify unexpectedly
high or low strength materials and materials that may exhibit brittle behaviour when
subjected to earthquake loading.
Note:
If the intention is to strengthen an existing steel building and the strengthening involves
welding to an existing steel, the weldability of the existing steel parent material also needs
to be determined.
where:
𝐻𝐻v = Vickers Hardness from test
𝑓𝑓u = probable tensile strength
This expression is valid for 100 ≤ 𝐻𝐻v ≤ 300, corresponding to 330 ≤ 𝑓𝑓u ≤ 950 MPa.
Note:
Testing for Vickers Hardness is carried out to AS 1817:1991 Metallic Materials – Vickers
Hardness Test (1991). There are a number of materials testing organisations in
New Zealand that can undertake Vickers Hardness tests.
The key steps for determining what components to test and how many tests to conduct are
as follows:
Step 1
Determine the members/elements to be tested, i.e. beams, columns, critical connection
components and connectors. The elements identified as critical from the connection
evaluation in Section C6.6.1 and the strength hierarchy evaluation in Section C6.7 should be
subjected to the most detailed testing.
Step 2
Determine the frequency of testing. The aim is to cover at least 5% of the total sample of
each type of critical component.
Step 3
Use Equation C6.1 or refer to Nashid et al. (2015) for the relationship between Vickers
Hardness and tensile strength.
Note:
Nashid et al. (2015) presents the findings of comprehensive recent research on the
hardness-tensile strength relationship of structural steel members.
Step 4
Compare the tensile strengths with the expected steel grades. Any material with 𝐻𝐻v < 100 or
𝐻𝐻v > 230 should be investigated more thoroughly by tensile sampling and visual inspection.
Any material with 𝐻𝐻v > 230 should also be treated as potentially prone to brittle fracture.
Note:
There is no direct relationship between tensile strength and brittle fracture. However, the
susceptibility to brittle fracture increases with increasing tensile strength. The elongation
capacity of steel also decreases with increasing strength. Accordingly, the guidance
provided above is a threshold requiring more appropriate testing for potential brittle
fracture performance.
• the Vickers Hardness test of the components identifies steel with 𝐻𝐻v > 230, or
• the thickness of any component is > 32 mm.
If any of these apply, material from those components should be removed for Charpy impact
tests, as specified in NZS 3404:1997, to determine whether the steelwork satisfies energy
absorption requirements. Test material may be removed from the less critical regions of a
member/element; e.g. from the web of beams away from high shear zones.
A minimum of three Charpy impact tests should be undertaken on material removed from
each type of critical component. For the energy absorption requirements to be satisfied, the
average Charpy impact energy absorption capacity of a steelwork from the three tests should
exceed 27 J at 0°C, while the minimum of the three tests should exceed 20 J at 0°C.
If the steel does not satisfy the above energy absorption requirements a more detailed
evaluation should be undertaken.
Note:
For brittle fracture of steel to occur during an earthquake, the steel has to have a low
Charpy impact energy absorption capacity at service temperature (or the steelwork has to
be below its transition temperature at the time of the earthquake) and a stress raiser has to
be present in a critical location.
Note:
In the absence of specific data, the probable strengths shown in Table C6.3 have been
taken as the lower characteristic strengths based on Bussell (1997) and ASCE 41-13
(2014) except for pre-1961 rivets which have been taken as 1.1 times their characteristic/
nominal strength.
Note:
*The probable yield strength of these rivets can be taken as half of their probable tensile strength.
The probable strength of structural steel members/elements should be determined using the
probable material strengths as outlined in Section C6.4. A strength reduction factor is not
required to be applied (i.e. a strength reduction factor, 𝜙𝜙, of 1.0 is used).
C6.5.2 Beams
C6.5.2.1 General
The probable strength of steel beams of seismic-resisting frames is generally governed by
flexural strength.
The flexural strength of a steel beam is dependent on the length of the beam between adjacent
cross sections that may be either restrained or unrestrained (segments) and the restraint
condition provided at the ends of the segments (full, partial or lateral restraint).
The effect of combined actions of shear and flexure should be assessed at cross sections
where both shear and flexure are expected to be high.
where:
𝑑𝑑p = depth of web
𝑡𝑡w = thickness of web
𝑓𝑓y = probable yield strength
the probable shear yield capacity of the web (𝑉𝑉v ) should be taken as (NZS 3404:1997):
where:
𝑓𝑓y = probable yield strength
𝐴𝐴w = area of web.
If the above web slenderness criterion is not satisfied and the web is slender, the web is likely
to buckle instead of yielding in shear. The probable shear buckling strength of slender webs
should be determined from Clause 5.11.5 of NZS 3404:1997.
The sections of a steel beam should be compact and not prone to local plate buckling in order
to have flexurally yielding regions in the beam that are able to develop and maintain their
full plastic section strength until the deformation capacity is reached.
In addition to having compact sections, steel beams or segments of steel beams need to have
full restraint against lateral buckling (FLR) to develop and maintain their full section plastic
strength.
Beams supporting a concrete slab are considered to have FLR and develop their section
flexural strength, while beams supporting timber floors generally achieve member flexural
strength only. Restraint offered to steel beams by timber floors or other lateral restraint
conditions are provided in HERA Report R4–92 (Clifton, 1997).
The probable strength of beams having slender sections is limited to their probable yield
strength due to local plate buckling. However, unlike elastic buckling of compression
members, buckling of slender plates of flexural members does not lead to immediate loss of
load-carrying capacity or excessive deflections, as shown in Figure C6.8, as redistribution
of in-plane stresses occurs within the plates.
A generic relationship between the probable flexural strength, 𝑀𝑀prob , and probable chord
rotation, 𝜃𝜃, capacity of steel beams with FLR is provided in Figure C6.9. The parameters for
the generic relationship for this type of beam should be taken from Table C6.4 using the
highest possible member category. The member category for steel beams should be
determined based on steel material and section geometry requirements outlined in
Clause 12.4 and Clause 12.5 respectively of NZS 3404:1997.
Figure C6.9: Moment-rotation relationship for steel beams and columns with FLR
Table C6.4: Parameters for the moment-rotation relationship for steel beams with FLR
1&2 45 0.5
3 30 0.5
Note:
If the beam under consideration cannot support gravity loading (𝐺𝐺 + 𝛹𝛹E 𝑄𝑄)
in a simply supported condition, halve the rotational capacity of the beam for
both the full and the residual strength capacity.
The probable yield rotation, 𝜃𝜃y , of seismic governed steel beams that are rigidly connected
to columns at both ends can be determined from:
𝑀𝑀prob,s 𝑙𝑙b
𝜃𝜃y = …C6.4
6𝐸𝐸𝐼𝐼b
where:
𝑀𝑀prob,s = probable section flexural strength
𝑙𝑙b = clear span of beam
𝐸𝐸 = modulus of elasticity
𝐼𝐼b = second moment of area of beam.
C6.5.3 Columns
Steel columns in seismic-resisting buildings are generally subjected to a combination of
flexure and axial forces. Both axial tension and compression reduce the flexural capacity of
steel columns, while axial compression reduces the local buckling capacity.
The probable strength of steel columns may be limited by the various member shear and
flexural yield mechanisms outlined in Section C6.5.2. In addition, the flexural capacity of
steel columns may be limited by column buckling. The probable section and member
flexural capacities of steel columns should be determined from Clause 8.3 and Clause 8.4 of
NZS 3404:1997 using probable material strengths.
When determining the rotation capacity of steel columns, the axial force used should be that
∗
from the gravity load associated with earthquake action (𝑁𝑁G+ψ EQ
) and the seismic
contribution should be ignored.
Note:
Experimental tests (MacRae, 1990 and Brownlee, 1994) have shown that the inelastic
behaviour and rotation capacity of a steel beam-column subject to compression and major
axis bending is dependent on the magnitude of the constant component of the compression
∗
force – i.e. that from 𝑁𝑁G+ψ EQ
– rather than on the total compression force that includes the
seismic component.
Steel columns that are subjected to inelastic demand should satisfy the axial load limitations
of Clause 12.8.3 of NZS 3404:1997. This clause is intended to ensure that the level of
compression in a column is not too high to compromise the capacity of the column to
dependably accommodate inelastic earthquake demands.
A typical moment-rotation relationship for steel columns that have FLR is provided in
Figure C6.9. The parameters for the generic relationship for this type of column should be
taken from Table C6.5 using the highest possible member category. The member category
for steel columns should be determined based on steel material and section geometry
requirements outlined in Clause 12.4 and Clause 12.5 respectively of NZS 3404:1997.
1&2 50 45 20 15 0.5
3 35 30 15 10 0.5
The probable yield rotation of steel columns that have a point of contraflexure at mid height
and are subjected to both flexure and compression can be determined from:
𝑀𝑀prob,s 𝑙𝑙c 𝑁𝑁 ∗
𝜃𝜃y = �1 − 𝑁𝑁 � …C6.5
6𝐸𝐸𝐼𝐼c prob,c
where:
𝑀𝑀prob,s = probable section flexural strength
𝑙𝑙c = clear length of column
𝐸𝐸 = modulus of elasticity
𝐼𝐼c = second moment of area of column
𝑁𝑁 ∗ = axial force from analysis
𝑁𝑁prob,c = probable member capacity in compression.
Comparisons of the provisions in NZS 3404:1997 with physical tests undertaken recently in
Canada (Clifton (not published at time of preparation)) on a medium heavy I-section column
type cross section and an I-section beam type cross section indicated that the following
modifications needed to be made to the provisions in NZS 3404:1997 when determining the
probable capacity of steel members:
• The rotation restraint factor (𝑘𝑘r ) should be taken as 0.85, consistent with a plastic hinge
forming at one end only at a particular point in time.
• The member effective length factor (𝑘𝑘e ) should be taken as 0.85 instead of the
NZS 3404:1997-specified 1.0, consistent with a plastic hinge forming at one end only at
a particular point in time.
The physical tests undertaken in Canada also showed that the moment-rotations parameters
presented in Table C6.5 for highly axially loaded columns are on the conservative side.
Members that are subjected to bending about their minor principal axis should be considered
capable of developing their probable plastic section flexural strength about their minor
principal axis.
The probable capacity of encased steel members not satisfying the requirements of
NZS 3404:1997 should be determined as discussed below.
Note:
The structural members of old steel frames are generally encased in lightly reinforced
concrete. In some cases the concrete encasement is unreinforced and has low compressive
strength, and is therefore generally considered to play a fire protection role only
(Bruneau and Bisson, 2000).
If the concrete encasement of old steel frames is reinforced, the reinforcement is often
nominal and consists of plain round bars and thin wire meshes. Inadequately reinforced
concrete encasement results in a significant increase in stiffness and a relatively small
increase in strength of the encased members.
where:
𝑆𝑆 = plastic section modulus
𝑓𝑓y = probable yield strength.
The moment-rotation relationship of concrete encased steel beams is similar to that provided
in Figure C6.9. The parameters for the generic relationship for this type of member should
be taken from Table C6.4 and the probable yield rotation from Equation C6.4.
The probable flexural capacity of such columns should be based on the probable flexural
capacity of the steel members only.
The probable flexural capacity of this type of column should be based on the probable
flexural capacity of the steel elements only, while the probable yield rotation may be taken
as 5 mrad in lieu of a detailed analysis.
The probable compression capacity of laced and battened columns should be determined
from Clause 6.4 of NZS 3404:1997 using probable material strengths.
C6.5.5 Braces
C6.5.5.1 Compression capacity
The performance of braces that are subjected to earthquake induced compression forces
principally depends on the slenderness ratio of the braces.
𝑘𝑘e 𝑙𝑙 𝑓𝑓y
Braces with a slenderness ratio � that is:
𝑟𝑟 250
> 120 generally do not have the capacity to carry compressive inelastic
earthquake demand and their capacity is exceeded typically through
elastic buckling
< 120 should be expected to buckle inelastically through local yielding
under the combined actions of compression and bending.
where:
𝑘𝑘𝑒𝑒 = member effective length factor
𝑙𝑙 = member length
𝑟𝑟 = radius of gyration
𝑓𝑓y = probable yield strength
When a compression brace buckles inelastically the same peak compression capacity as
achieved in a previous cycle is generally not likely to be achieved during subsequent cycles
of loading. A typical force-displacement relationship for a brace in compression is presented
in Figure C6.10. The values of the parameters in the figure are provided in Table C6.6.
Note:
∗ given by Equation C6.7 or C6.8 as appropriate
𝑘𝑘e 𝑙𝑙 𝑓𝑓y
𝜆𝜆n should be determined from Clause 6.3.3 of NZS 3404:1997 as: 𝜆𝜆n = �
𝑟𝑟
� �𝑘𝑘f �
250
∆c is the probable elastic axial deformation of a brace at buckling (𝑁𝑁 ∗ =𝑁𝑁prob,c )
∆p is the probable plastic axial deformation capability of a brace before degradation of strength
∆cap is the probable deformation capacity of a brace.
The residual strength factor (α′c ) for compression braces that are likely to remain nominally
ductile due to XXX%ULS shaking is given as:
α′c = 7.7⁄𝜆𝜆0.6
n ≤ 1.0 …C6.7
and for compression braces that are likely to be subjected to ductile demand at XXX%ULS
shaking as:
α′c = 42.15⁄𝜆𝜆1.1
n ≤ 1.0 …C6.8
Braces in tension are not considered to have any residual capacity after their displacement
capacity, Δcap , is exceeded.
Note:
The probable tensile capacity of a brace should not be taken as greater than the probable
capacity (in tension and/or in shear) of the connections at either end.
I-section 11∆t
Note:
∆t is the axial deformation of a brace at yield (𝑁𝑁 ∗ =𝑁𝑁prob,s ).
The probable shear and flexural capacities of an active link should be determined from
Section C6.5.2.
The force-rotation relationship of active links is similar to that provided in Figure C6.11.
However, 𝑁𝑁prob,s , Δy and Δcap in Figure C6.11 should be replaced with 𝑉𝑉𝑤𝑤 , 𝛾𝛾y and 𝛾𝛾cap
respectively to obtain the force-rotation relationship for active links. The values of the
parameters in the force-rotation relationship are provided in Table C6.8.
Note:
𝛾𝛾y is the rotational deformation of an active link at yielding.
The following advice should help when determining the load path through a connection and
the weakest link in a load path:
• Determine the internal forces that could be generated in the attached members during an
earthquake.
- An I-section beam responding elastically under flexure will deliver axial forces
through the flanges (tension and compression) and vertical shear through the web.
- An I-section beam responding inelastically under flexure will deliver axial yield
forces through the flanges and axial yield forces plus vertical shear through the web.
- A brace will deliver axial forces (tension is critical) through all its elements.
• Trace the transfer of forces from elements of the supported member into elements of the
supporting member that lie parallel to the incoming force. For example, the incoming
axial forces from an I-section beam flange connected to an I-section column should be
transferred through the column flange into the column web.
• Calculate the probable capacity of all elements along the identified load path in
accordance with the provisions of Sections C6.6.3 and C6.6.4.
• If there are no tension and compression stiffeners in a column adjacent to incoming beam
flanges in a moment-resisting beam to column connection, then tensile distortion of the
flange of the column or compression buckling of the web of the column web are likely
to occur before the beam can develop its full flexural capacity.
• The strength and ductility capacity of a load path is determined by the strength and
ductility capacity of the weakest component in the load path.
• If various load paths exist through a connection, the stiffest of the load paths will attract
the most force.
• Be particularly aware of situations where the connectors (rivets, bolts or welds) may be
the weakest component, as their ductility capacity will be limited. One sided fillet welds
in tension or bending are particularly vulnerable in this regard, showing no ductility.
• Be aware of component forces introduced when an applied force changes direction along
the load path.
Note:
The article by Blodgett (1987) on welds explains the concept of load paths through welded
connections and illustrates this with a number of examples.
SP GP
Incomplete penetration butt, fillet, plug and slot welds 0.9 0.7
Note:
Strength modification coefficients are not to be confused with strength reduction factors,
which for assessment are taken as 1.0. Strength modification coefficients are intended to
better define the probable capacity of the defined components.
The assessment of bolted and riveted connections should be undertaken in accordance with
the following:
• Probable shear capacity of rivets, 𝑉𝑉prob,f , can be determined from Barker (2000). The
key equation is derived from the bolt shear capacity provisions of NZS 3404:1997 and
is given as:
where:
𝑓𝑓uf = probable tensile strength of the rivet
𝑘𝑘r = reduction factor given in Table 9.3.2.1 of NZS 3404:1997 to
account for the length of a lap connection (𝐿𝐿j ). 𝑘𝑘r = 1.0 for
𝐿𝐿j < 300 mm and for all other type of connections
𝑛𝑛x = number of connector shear planes intercepting the shear plane
𝐴𝐴o = nominal plain shank area of the rivet.
where:
𝐴𝐴s = gross tensile stress area of the rivet.
𝑓𝑓uf = probable tensile strength of the rivet
• The diameter of a rivet shank should be determined from the diameter of the rivet head
in accordance with Figure C6.13.
• Be aware that some less scrupulous erectors made up some dummy rivets from moulded
putty covered in paint on larger groups of rivets. Hitting each rivet with a hammer will
soon identify any dummy ones!
• Assume that concrete encasement, if present and with any amount of confining
reinforcement, will prevent local buckling of the steel members. This assumption may
not hold for members in regions subject to significant inelastic demand and will need to
be assessed more closely for such regions.
Figure C6.13: Typical rivet shank and head diameters (Bussell, 1997)
The experimental tests undertaken on historical riveted connections by Roeder et al. (1996)
revealed that the mode of failure of clip angle connections under cyclic loading was similar
to that under monotonic loading. Both monotonic and cyclic load tests deteriorate and fail at
similar levels deformation demands, as shown in Figure C6.14. The monotonic tests
typically provided an upper bound envelope for the cyclic tests.
Both concrete encased and bare connections were observed to experience strength
degradation at rotations in the order of 20-25 milliradians. It was also observed that concrete
encasement improved performance by suppressing any local deformation until the concrete
was crushed at larger deformation demands due to lack of adequate confinement.
It should be noted that bolted clip angle connections would be stiffer and would have more
rotational capacity than comparable riveted connections. However, the limits on the overall
system inelastic displacement would be such that bolted connections cannot attain their full
capacity. For example, when the connections are the weakest element, rotational demand on
the connections will be around 30 milliradians maximum for an inter-storey drift of 2.5%.
Therefore, a 40 milliradians limit on rotation is considered a practical upper limit for the
system as a whole, even if the individual connection is capable of greater rotations while
maintaining a dependable level of flexural capacity.
Figure C6.15: Comparison of bare steel and encased moment–rotation behaviour of clip
angle connections (Roeder et al., 1994)
General
The strength and rotation capacity of bolted and riveted clip angle connections (illustrated
in Figure C6.16) can be determined from first principles and using the guidance presented
in this section. The procedure includes a method for determining the probable flexural
strength, along with expressions for estimating the probable rotational capacity. Both
flexural strength and degradation threshold are considered to be a function of the expected
mode of failure of the connections to the beam flanges.
t1 Bf
m e
P a1
a X
T
t2 Db
b k
C
2 x la
Dc Bc
The flexural strength of a clip angle connection is limited by the smallest demand required
to form one of the following four yielding/shear failure modes (Roeder et al., 1996):
• shear yielding/failure of the connectors
• tensile capacity of flange cleat angles
• tensile capacity of connectors, or
• flexural yielding of connection elements (flange cleat angles and/or web cleat angles).
The probable shear strength of connectors, 𝑉𝑉prob,f, can be determined from Equation C6.9.
The probable flexural strength of a clip angle connection limited by the shear strength of the
connectors, 𝑀𝑀prob , can be determined from:
where:
𝑛𝑛 = the number of connectors
𝐷𝐷b = the depth of the beam.
where:
𝐴𝐴g = gross area of the cross section
𝑓𝑓y = probable yield strength of the section
𝑘𝑘te = tcorrection factor in accordance with Clause 7.3 of NZS 3404:1997
𝐴𝐴n = net area of the cross section
𝑓𝑓u = probable tensile strength of the section.
The probable flexural strength of a clip angle connection limited by the tension capacity of
the flange angles, 𝑀𝑀prob , can be determined from:
𝑡𝑡
𝑀𝑀prob = 𝑁𝑁prob,t �𝐷𝐷b + 1�2� …C6.13
where:
𝑡𝑡1 = the thickness of the flange cleat angle leg.
where:
𝑛𝑛 = the number of connectors
𝑎𝑎 = the distance between the centre of the connectors and the flange cleat
angle leg.
Flexural yielding of the flange cleat angle requires development of prying actions. However,
the prying forces that develop in connections that use mild steel connectors are typically not
likely to cause the capacity of the connectors to be exceeded.
The probable flexural strength of a clip angle connection reduced by prying actions, 𝑀𝑀prob ,
is given as:
where:
𝐵𝐵f = the length of the angle
𝑎𝑎1 = the distance between the centre of the connectors and the top edge of
the flange cleat angle.
If the connectors are strong enough to induce flexural yielding of the flange cleat angles, the
probable flexural strength can be determined from:
𝐵𝐵f 𝑡𝑡12
𝑀𝑀prob = 𝑓𝑓y (𝐷𝐷b + 𝑡𝑡1 ⁄2) …C6.16
2𝑎𝑎
The probable flexural capacity of the web cleat angle can be determined from:
𝑙𝑙a 𝑡𝑡22
𝑀𝑀prob = 𝑓𝑓y …C6.17
2
where:
𝑙𝑙a = the length of the web cleat angle face
𝑡𝑡2 = thickness of the web cleat angle leg
𝑓𝑓y = probable yield strength.
From Equation C6.17, the tensile force in the web cleat bolts/rivets is:
2𝑀𝑀prob
𝑇𝑇 = …C6.18
𝑘𝑘
where:
𝑘𝑘 = the distance between bolt centreline and the web cleat angle leg.
where:
𝑚𝑚 = distance from centre of bolt hole to radius root at web
𝑒𝑒 = distance from rivet centre to flange edge
𝑡𝑡c = thickness of the column flange
𝑓𝑓yc = probable yield strength of the column flange.
The contribution of the web cleat angle to the probable flexural strength of the connection
is:
where:
𝑄𝑄 = either T from Equation C6.18 or 𝑇𝑇c from Equation C6.19, whichever
is less
𝑏𝑏 = the distance between the centroid of tension and compression forces
in the web cleat.
Note:
𝑑𝑑b is depth of beam (m)
𝑀𝑀prob,encased = 2𝑀𝑀prob,bare
Note:
As discussed in Section C6.3.2, fractures of welded beam-column connections were
widely reported after the 1994 Northridge earthquake, with the majority of these fractures
observed at the bottom beam-column flange connections. Refer to that section for more
discussion, including a list of the factors considered to have contributed to the brittle
failures of “Pre-Northridge” connections (FEMA 355E, 2000).
b Flange weld
f Flange weld
Beam flange capacity 𝑁𝑁fbt = 𝑁𝑁fbc = 1.2[min(𝑏𝑏fb , 𝑏𝑏fc )]𝑡𝑡fb 𝑓𝑓yb …C6.21
Column web tension capacity 𝑁𝑁wct = [𝑡𝑡fb + 2𝑠𝑠f + 5(𝑡𝑡fc + 𝑟𝑟c )]𝑡𝑡wc 𝑓𝑓yc …C6.23
2
𝜂𝜂 = �(1.15 − (𝑁𝑁 ∗ ⁄𝑁𝑁s ) ≤ 1
Note:
*If 𝑏𝑏eff < 0.7𝑏𝑏fb , tension stiffeners are necessary to avoid weld tearing at the point of peak stress.
𝑏𝑏1 , 𝑛𝑛1 , & 𝑛𝑛2 should be reduced if the column projection is insufficient for full dispersal.
+ If doubler plates are provided, (𝑡𝑡wc + 𝑡𝑡p ) should replace 𝑡𝑡wc in Equation C6.26.
where:
𝑏𝑏1 = width of contact between beam flange and welds and column
𝑏𝑏eff = effective beam flange width
𝑏𝑏fb = beam flange width
𝑏𝑏fc = column flange width
𝑑𝑑b = beam depth
𝑑𝑑c = column depth
𝑓𝑓yb = probable yield strength of beam
𝑓𝑓yc = probable yield strength of column
𝑓𝑓uw = probable strength of weld metal
𝑛𝑛1 = length obtained by a 45° dispersion though half the depth of the column
𝑛𝑛2 = length obtained by a 1:2.5 dispersion though column flange and root radius
𝑁𝑁 ∗ = axial load in column below joint
𝑁𝑁fbc = probable compression capacity of beam flange
𝑁𝑁fbt = probable tension capacity of beam flange
𝑁𝑁fct = probable tension capacity of column flange
𝑁𝑁s = probable column section compression capacity
𝑁𝑁tfw = probable tension capacity of beam flange weld
𝑁𝑁wcc = probable compression capacity of column web
𝑁𝑁wct = probable tension capacity of column web
𝑟𝑟c = column root radius
𝑠𝑠f = weld leg length to beam tension flange (when available)
𝑡𝑡fb = beam flange thickness
𝑡𝑡fc = column flange thickness
𝑡𝑡p = total thickness of doubler plates
𝑡𝑡wc = column web thickness
𝑉𝑉c = probable shear capacity of panel zone.
∗ 𝑀𝑀∗ 𝑁𝑁 ∗
𝑁𝑁fbt = 𝑑𝑑 − …C6.28
b −𝑡𝑡fb 2
∗ 𝑀𝑀∗ 𝑁𝑁 ∗
𝑁𝑁fbc = 𝑑𝑑 + …C6.29
b −𝑡𝑡fb 2
where:
∗
𝑁𝑁fbt = tension demand on beam flange
∗
𝑁𝑁fbc = compression demand on beam flange
𝑁𝑁 ∗ = axial load in column below joint
𝑀𝑀∗ = moment in beam.
If the various components of a welded connection do not have the required capacity to resist
beam/column overstrength demand, as would be the case for an unstiffened column that is
typical of old buildings, the moment-rotation behaviour of the connection should be taken
from Table C6.13 and the general shape of the moment-rotation curve should take the form
of Figure C6.17.
𝑀𝑀prob,w is the probable flexural capacity of the beam web column connection and needs to
be determined from the particular connection detail adopted. This capacity is determined
from:
• the probable capacity of the connection, if the beam web is connected to the column
flange using clip angles, or
• the probable plastic flexural capacity of the beam web, if the beam web is connected
using balanced, double sided fillet welds, or butt welds of sufficient strength to yield the
web in tension.
If a beam-column connection is suspected of being welded but the connection is not visible
(e.g. due to concrete encasement) and if no drawings are available, the encasement material
should be removed from a representative connection so that a reasonable assessment can be
undertaken. The difference in connection moment-rotation capacity between a connection
that can transfer the beam flange axial forces induced by inelastic beam action dependably
into the column and one that cannot is significantly large that the capacity should be
determined and not guessed.
Similarly, the existing state of welds needs to be assessed using visual inspection techniques.
Engineers undertaking weld inspections should be familiar with visual inspection techniques
such as those outlined by Hayward and McClintock (1999).
If the concrete encasement is well reinforced and is likely to contribute to the strength and
stiffness of the steel frame, the contribution of the composite section should be determined.
Note that this is very unlikely in pre-1976 building encased beams and is more likely in pre-
1976 building encased columns. Column encasement is advantageous as it increases column
strength relative to beam strength.
• The assessed inelastic response of the system (this assessment is qualitative rather than
quantitative) is essentially symmetrical in nature and does not contain features that will
inevitably lead to a progressive deformation of the building in one direction only.
If the ductility demand on a steel frame due to XXX%ULS shaking is not significant
(𝜇𝜇 ≤ 1.5) and the frame exhibits the four “good features” listed above, the inelastic response
of the frame does not need to be assessed.
Step 1
Determine the probable material strength of the members, the elements of the connections
and the connectors. Use probable strengths provided in Section C6.4 in the absence of
original construction documentation and physical test data.
Step 2
Determine and assemble the probable capacity of the individual members and connections
located on potentially critical floor levels. Refer to Sections C6.5.2 and C6.5.3 for beams
and columns respectively, and Section C6.6 for connections.
If the individual beams of the frame on each level under consideration cannot support gravity
loading (𝐺𝐺 + 𝜓𝜓E 𝑄𝑄) in a simply supported condition, then halve the plastic rotation capacity
of the beams (refer to Section C6.5.2) and of the connections (refer to Section C6.6).
The reduction in rotational capacity reflects the monotonic, cumulative nature of inelastic
demand on the yielding regions of such members.
If the slab is placed in contact with the columns of a frame or insufficient separation is
provided, the contribution of the slab to the flexural strength of the beams should be taken
into account.
The assessment should include the first level above the seismic ground level, the uppermost
seismic mass level, and floor levels where member sizes and/or connection types change.
Step 3
Determine the governing inelastic mechanism of the frame: i.e. beam sidesway or column
sideway mechanism.
A sway potential index (𝑆𝑆i ) can be employed to determine the potential sway mechanism of
a frame. A sway potential index can be defined at a storey of the frame by comparing the
sum of the probable flexural strengths of the beams (or connections, whichever are smaller)
and the columns at the centroid of every joint:
∑(𝑀𝑀 +𝑀𝑀 )
𝑆𝑆i = ∑(𝑀𝑀bl +𝑀𝑀br ) …C6.30
ca cb
where:
𝑀𝑀bl = probable beam (or connection, whichever is smaller) flexural strength
to the left of the joint, extrapolated to the centroid of the connection
𝑀𝑀br = probable beam flexural strength (or connection, whichever is smaller)
to the right of the joint, extrapolated to the centroid of the connection
𝑀𝑀ca = probable column flexural strength above the joint, extrapolated to the
centroid of the connection
𝑀𝑀cb = probable column flexural strength below the joint, extrapolated to the
centroid of the connection.
If:
• 𝑆𝑆i < 0.85, a beam sidesway mechanism is likely to form. It should be noted that a
significant change in storey heights increases the likelihood of a column sidesway
mechanism.
• 0.85 < 𝑆𝑆i < 1, either a beam sidesway or column sidesway mechanism is likely to form.
The effect of both mechanisms need to be assessed.
• 𝑆𝑆i > 1, a column sidesway mechanism is likely to form.
Note:
When a frame has semi-rigid connections and these connections are flexurally weaker
than the beams or the columns, a beam sidesway mechanism forms.
Step 4
Determine the probable base shear capacity, 𝑉𝑉prob , of the frame.
If the potential inelastic mechanism of a frame is beam sidesway, the probable base shear
capacity, 𝑉𝑉prob , of the frame can be determined from (refer to Figure C6.19):
∑m n
i=1 𝑀𝑀ri + ∑i=1 𝑉𝑉bi 𝐿𝐿eq
𝑉𝑉prob = …C6.31
ℎeq
where:
𝑀𝑀prob,I = probable flexural strength of column i at the base
𝑉𝑉bi = storey i beam seismic shear demands determined from
probable beam flexural strengths as:
𝑀𝑀prob,il + 𝑀𝑀prob,ir
𝑉𝑉bi = …C6.32
𝐿𝐿b
Note:
Equation C6.31 provides an upper bound base shear capacity. For the frame to achieve
this upper bound base shear capacity, all of the beams and the bases of the columns should
start to yield before the rotational capacity of the critical hinge is exceeded.
If the rotational capacity of the critical hinge is likely to be exceeded before some of the
beams and/or the bases of the columns start to yield, the flexural resistance developed in
the potential plastic hinges that have not started to yield should replace probable flexural
strengths in Equation C6.32.
𝑉𝑉prob = 𝑉𝑉base
If the potential inelastic mechanism of a frame is column sidesway, the probable base shear
capacity, 𝑉𝑉prob , of the frame can be estimated from:
∑m m
i=1 𝑀𝑀ri,b +∑i=1 𝑀𝑀ri,t
𝑉𝑉prob = …C6.34
ℎ
where:
𝑀𝑀prob,ib = probable column flexural strengths at the base or bottom of
column i extrapolated to the centroid of the connection
𝑀𝑀prob,it = probable column flexural strengths at the top of column i
extrapolated to the centroid of the connection
ℎ = storey height.
Step 5
Ensure the axial force demand on the external columns does not exceed the probable limiting
axial force (𝑁𝑁cr ) from Section 12.8.3.1 of NZS 3404:1997:
∗ ∗
𝑁𝑁eq + 𝑁𝑁G+ψ𝐸𝐸 Q
≤ 𝑁𝑁cr …C6.35
where:
∗
𝑁𝑁eq = earthquake-induced axial force demand
∗
𝑁𝑁G+ψ𝐸𝐸 Q = axial force demand due to gravity loading.
If the above equation is not satisfied, reduce the base shear capacity of the frame until it is.
Step 6
Determine the deformation capacity of the frame.
The assessment of infilled steel MRFs should allow for the stiffening effect of infill panels
on the overall system response. In addition, the presence of infill panels induces increased
shear demands on the frame members by creating short column effects. The increased shear
demands are unlikely to exceed the capacity of bare steel or concrete encased solid section
columns. However, elements of encased laced and battened members may not have
sufficient shear capacity. In addition, if the columns have a better shear capacity than the
infills and the infills are likely to sustain significant damage, the potential for a soft-storey
formation should be taken into consideration.
Steel moment-resisting infilled frames with weak connections should be assessed for the
potential for diagonal compression struts formed in infill panels pulling apart beam to
column connections as the frames deform laterally. External beam to column connections
are likely to be more critical than internal connections.
Note:
The assessment of weak beam to column connections involves comparison of the tension
capacity of the connections with the peak compression capacity of the infill panels
(capacity prior to deterioration due to panel crushing/shear failure). If the infill panel
compression strut capacity is greater than the beam to external column connection tension
capacity, failure of this connection needs to be considered for the response of that end bay.
Note:
When a compression brace of a V-braced CBF buckles, the capacity of the tension brace
may not be fully utilised as the collector beam may not have the capacity to resist the
unbalanced vertical force acting at the brace-collector beam joint. Note that the collector
beam will have to resist demands due to gravity loads in addition to the unbalanced vertical
force.
The buckling of compression braces of V-braced CBFs results in significant reductions in
frame lateral stiffness and strength, as the system changes to a D-braced EBF with a long
flexural link. In such situations a plastic hinge is likely to form in the collector beam before
the tension brace yields in tension.
During the subsequent reversing cycle of earthquake demand, the previously tension brace
generally buckles before the braces that buckled during the preceding half cycle fully
straighten up (Tremblay and Robert, 2000). Therefore, the inelastic capacity of V-braced
frames is limited by the post-buckling capacity of the braces.
The following steps outline an assessment procedure for X-braced and V-braced CBFs.
Step 1
Determine and assemble the probable capacity of the individual members and connections
located on potentially critical floor levels. The capacities to be determined are:
• axial force capacity of the braces
• post-buckling capacity of the braces
• flexural and compression capacity of collector beams
• axial force and flexural capacity of columns
• axial force capacity of connections and splices.
Step 2
Determine the weakest member and the expected mode of failure; i.e. brace, brace
connection, collector beam, column, etc.
Step 3
Step 4
If the ductility demand on the frame due to XXX%ULS shaking is not significant (𝜇𝜇 ≤ 1.5)
and the frame exhibits the above “good features”, the inelastic response of the frame need
not be assessed.
If the braces or brace connections are not the weakest component, the capacity of the frames
should be limited to the capacity of the weakest member/element if the failure of that
member/element constitutes loss of gravity load carrying capacity.
If the brace connections are the weakest component resulting in a rather low lateral force
capacity, the frame can be assessed as a moment resisting frame. However, the failure of the
brace connections is unlikely to lead to loss of gravity load carrying capacity on their own.
If the ductility demand on the frame due to 100%ULS shaking is significant (𝜇𝜇 > 1.5) and
the braces are the weakest component, proceed to the next step.
Step 5
Determine the probable base shear capacity of the frame.
The capacity of CBFs can be determined from first principles and the brace member capacity
relationships provided in Section C6.5.5.
𝑉𝑉prob = ∑m ′
i=1 𝛼𝛼ci 𝑁𝑁prob,ci 𝑐𝑐𝑐𝑐𝑐𝑐 𝜃𝜃i …C6.36
where:
′
𝛼𝛼ci = residual strength factor for brace i from Section C6.5.5.1
𝑁𝑁prob,ci = probable compression capacity of brace i
𝜃𝜃i = angle between brace i and beam at the top end of the brace
m = number of braces.
If a soft storey forms in one of the upper storeys of a CBF, the calculated base shear capacity
should allow for the resistance mobilised in the braces that are located in the storeys below
the soft-storey level before the capacity of the critical brace is exceeded.
Step 6
Determine the displacement capacity of the frame.
The yield displacement, Δy , of a CBF may be determined from an elastic analysis of the
frame based on the displacements at which the first brace yields in tension (if a tension-only
brace) or buckles in compression and a mechanism develops in a storey.
where:
𝐿𝐿b = length of the critical brace
∆cap,b = displacement capacity of the critical brace from
Section C6.5.5.1
ℎ = storey height
𝐿𝐿 = width of the braced bay.
The global assessment of steel framed buildings may be undertaken using either a
displacement or force based assessment procedure as appropriate. This section covers factors
specific to the analysis and assessment of these buildings. Refer to Section C2 for overall
procedures and appropriate global analysis methods.
Rotational stiffness of column base connections can be determined from NZS 3404:1997 as:
𝑘𝑘𝑘𝑘𝐼𝐼c
𝑘𝑘θ = …C6.38
𝐿𝐿c
where:
𝑘𝑘 = 1.67 for fixed base connections
𝑘𝑘 = 0.1 for pinned base connections
𝐼𝐼c = second moment of area of the column about the direction under
consideration
𝐿𝐿c = length of column.
𝐸𝐸 = modulus of elasticity
Note:
Experimental tests undertaken on typical seismic-resisting system foundations have
confirmed that the fixed base rotational stiffness recommendation of NZS 3404:1997 is a
reasonable value to adopt (AISC, 2012; Borzouie et al., 2016).
When undertaking an elastic analysis of a steel framed building, rigid end blocks having
dimensions equal to one half of the beam depth and one half of the column depth should be
used at each beam to column connection of the lateral force resisting frame. The use of one
half the depth as a rigid end block instead of the full member depth takes account of the
flexibility of the panel zone of the connections.
Comparison of the probable strength of the assumed non-yielding members and connections
with the actions generated by the overstrength of potential plastic hinge regions determines
whether the members and connections outside these hinge regions of the building are
protected. If they are not, development of the full ductile mechanism may not be possible
and the overall capacity of the mechanism may reduce.
To ensure that a ductile mechanism can develop as assumed, the assumed non-yielding
members and connections should have a probable capacity that is greater than required to
resist the actions resulting from yielding the plastic hinge regions at overstrength.
In order to meet the objectives of these guidelines the overstrength actions should be
determined using the maximum overstrength factor defined for the particular steel grade in
Table 12.2.8(1) of NZS 3404:1997 irrespective of the Category designation.
An X-braced roof diaphragm that remains close to elastic is likely to fulfil its role more
reliably than one that yields (noting that there will be a level of earthquake that will cause
actions beyond elastic levels unless the actions are limited by a reliable mechanism in the
lateral force resisting system supporting the roof). It should be also recognised that a
diaphragm that can reliably yield will likely perform better than one that cannot.
Maximum actions on an X-braced roof diaphragm are principally dependent on the capacity
of the lateral force resisting system supporting the roof and the capacity of the connections
of the roof bracing system.
References
AISC (2012). Column base connections, www.aisc.org/content.aspx?id=31382, American Institute of Steel
Construction, Chicago, USA.
AS 1163-1973. Structural steel hollow sections, Standards Association of Australia, North Sydney, New South
Wales, Australia.
AS 1163-1981. Structural steel hollow sections, Standards Association of Australia, North Sydney, New South
Wales, Australia.
AS 1163-1981 (Amdt 2). Structural steel hollow sections, Standards Association of Australia, North Sydney,
New South Wales, Australia.
AS 1163-1991. Structural steel hollow sections, Standards Australia, Homebush, New South Wales, Australia.
AS 1204-1972. Structural steels – Ordinary weldable grades, Standards Association of Australia, North Sydney,
New South Wales, Australia.
AS 1204-1980. Structural steels – Ordinary weldable grades, Standards Association of Australia, North Sydney,
New South Wales, Australia.
AS 1205-1980. Structural steels - Weather-resistant weldable grades, Standards Association of Australia, North
Sydney, New South Wales, Australia.
AS 1405-1973. Carbon steel plates of structural quality, Standards Association of Australia, Sydney, Australia.
AS 1817:1991. Metallic materials – Vickers hardness test, Standards Australia, Sydney, Australia.
AS 3678-1990. Structural steel – Hot-rolled plates, floorplates and slabs, Standards Association of Australia,
North Sydney, New South Wales, Australia.
AS 3679-1990. Hot-rolled structural steel bars and sections, Standards Association of Australia, North Sydney,
New South Wales, Australia.
AS A1-1928. Structural steel and Australian Standard rolled steel sections for structural purposes, Australian
Commonwealth Engineering Standards Association, Sydney, Australia.
AS A1-1956. Structural steel (excluding plates) and Australian Standard rolled steel sections for structural
purposes, Standards Association of Australia, Sydney, Australia.
AS A135-1965. Notch ductile steel for general structural purposes, Standards Association of Australia, Sydney,
Australia.
AS A149-1965. Mild steel for general structural purposes, Standards Association of Australia, Sydney, Australia.
AS A151-1966. Structural steel of high yield stress (welding quality), Standards Association of Australia, Sydney,
Australia.
AS A157-1966. Low and intermediate tensile strength carbon steel plates of structural quality, Standards
Association of Australia, Sydney, Australia.
AS A186-1971. Structural steels ordinary weldable grades, Standards Association of Australia, Sydney,
Australia.
AS A187-1971. Structural steels (weather-resistant weldable grades), Standards Association of Australia,
Sydney, Australia.
AS A33-1937. Carbon steel plates for general structural engineering purposes, Standards Association of
Australia, Sydney, Australia.
AS A33-1955. Carbon steel plates for general structural engineering purposes, Standards Association of
Australia, Sydney, Australia.
AS/NZS 1163:2009. Cold-formed structural steel hollow sections, Standards Australia/Standards New Zealand.
AS/NZS 3678:1996. Structural steel – Hot-rolled plates, floorplates and slabs, Standards Australia/Standards
New Zealand.
AS/NZS 3678-2011. Structural steel—Hot-rolled plates, floorplates and slabs, Standards Australia/Standards
New Zealand.
AS/NZS 3679.1:1996. Hot-rolled bars and sections, Standards Australia/Standards New Zealand.
AS/NZS 3679.1:2010. Hot-rolled bars and sections, Standards Australia/Standards New Zealand.
ASCE 41-13 (2014). Seismic evaluation of existing buildings, American Society of Civil Engineers, and
Structural Engineering Institute, Reston, Virginia, USA.
ASM International (1986). Guide to materials engineering data and information, ASM International, Ohio, USA.
Baker, M.J. (1969). Variations in the mechanical properties of structural steel - British early 1960s, IABSE reports
of the working commissions. No. 4. 165-174.
Barker, G.F. (2000). Assessment of existing structures: a strength limit state appraisal of hot driven rivets, Steel
Construction 34(1): 1–15, March 2000.
Bates, W. (1991). Historical Structural Steelwork Handbook. 4th ed. The British Constructional Steelwork
Association Limited.
Blodgett, O.W. (1987). Weld failures: they could be the result of violating simple design principles, Australia
Welding Journal 32(2).
Borzouie, J., MacRae, G.A., Chase, J.G., Rodgers, G.W. and Clifton, G.C. (2016). Experimental studies on
cyclic performance of column base strong axis aligned asymmetric friction connections, Journal of Structural
Engineering 142(1).
Brownlee, S.A. (1994). Axial force and plate slenderness effects on the inelastic behaviour of structural steel
beam-columns, University of Canterbury, Christchurch, NZ.
Bruneau, M. and Bisson, M. (2000). Hysteretic behavior of existing and retrofitted concrete-encased riveted
stiffened seat angle connections, Engineering Structures 22(9): 1086-1096.
BS 15:1948. Structural steel, British Standards Institution, London, UK.
BS 15:1961. Mild steel for general structural purposes, British Standards Institution, London, UK.
BS 4360:1968. Weldable structural steels, British Standards Institution, London, UK.
BS 548:1934. High tensile structural steel for bridges and general building construction, British Standards
Institution, London, UK
BS 968:1943. War time amendment No. 1, (War emergency standard) High tensile (fusion welding quality)
structural steel for bridges and general building purposes, British Standards Institution, London, UK.
Bussell, M. (1997). Appraisal of existing iron and steel structures, The Steel Construction Institute, Ascot,
England, SCI Publication 138.
Butcher, G., Andrews, L. and Cleland, G. (1998). The Edgecumbe earthquake, a review of the 2 March 1987
Eastern Bay of Plenty earthquake, Centre for Advanced Engineering, University of Canterbury, Christchurch,
NZ.
Butterworth, J.W. (1995). Inelastic rotation capacity of Grade 300 and Grade 350 universal columns, Department
of Civil and Resource Engineering, University of Auckland. School of Engineering Report No. 554.
Clifton, G.C. (1996). Steel building performance in two recent major earthquakes, IIW Asian Pacific Welding
Congress, Auckland, NZ.
Clifton, G.C. (1997). Restraint classifications for beam member moment capacity determination to
NZS 3404:1997, HERA, Manukau City, HERA Report R4–92.
Clifton, G.C. (2009). Lateral restraint of yielding regions in columns and beams in multi-storey buildings, Steel
Construction New Zealand. Steel Advisor MEM2001.
Clifton, G.C. and Cowie, K. (2013). Seismic design of eccentrically braced frames, HERA Publication
P4001:2013.
Clifton, G.C. and El Sarraf, R. (2011). New Zealand Steelwork Corrosion Coatings Guide. HERA, Manukau City,
HERA Report R4–133.
Clifton, G.C. and Ferguson, W.G. (2015). Determination of the post-earthquake capacity of an eccentrically
braced frame seismic resisting system, The University of Auckland, Report to the Natural Hazards Research
Platform.
Erasmus, L.A. (1984). The mechanical properties of structural steel sections and the relevance of these
properties to the capacity design of structures, Transactions Vol. II, No. 3/CE.
Erasmus, L.A. and Smaill, J.S. (1990). The mechanical properties of BHP structural sections, Transactions
Vol. 17, No. 1/CE, No. 1/EMCh.
Feeney, M.J. and Clifton, G.C. (2001). Seismic design procedures for steel structures, HERA Report R4-76,
Manukau City, New Zealand. HERA, 1995. To be read with Clifton, G.C.; Tips on seismic design of steel
structures, Notes from Presentations to Structural Groups mid-2000; HERA, Manukau City, 2000.
FEMA 273 (1997). NEHRP guidelines for the seismic rehabilitation of buildings, Federal Emergency
Management Agency, FEMA Report 273, Washington, DC.
FEMA 355E (2000). State of the art report on past performance of steel moment-frame buildings in earthquakes,
Federal Emergency Management Agency, FEMA Report 355E, Washington, DC.
FEMA 356 (2000). Prestandard and commentary for the seismic rehabilitation of buildings, Federal Emergency
Management Agency, FEMA Report 356, Washington, DC.
Ferris, H.W. (1954). Rolled Shapes – Beams and columns – period 1873 to 1952, American Institute of Steel
Construction, New York, USA.
Fidler, H. (1879). Notes on building construction, Vol. 3, Rivingtons.
Gupta, A. and Krawinkler, H. (1999). Seismic demands for performance evaluation of steel moment resisting
frame structures, The John A. Blume Earthquake Engineering Center, Report No. 132.
Hayward, P. and McClintock, A. (1999). What every engineer should know about welding and inspection,
CBIP/NZWC, HERA, Manukau City.
Hyland, C. (1999). Structural steelwork connections guide, HERA, Manukau City, HERA Report R4–100.
Kotwal, S. (2000). Evolution of Australian Standards for structural steel, 1st ed. Standards Australia International.
Krawinkler, H. (1995). Earthquake design and performance of steel structures, Pacific Conference on
Earthquake Engineering.
MacRae, G.A. (1990). The seismic response of steel frames, University of Canterbury Research Report 90–6.
University of Canterbury, Christchurch, NZ.
Nashid, H., Clifton, G.C., Ferguson, G and Choi. J. (2015). Relationship between hardness and plastically
deformed structural steel elements, Earthquakes and Structures, Vol. 8, No. 3.
Norton, J.A., King, A.B., Bull, D.K., Chapman, H.E., McVerry, G.H., Larkin, T.J. and Spring, K.C. (1994).
Northridge earthquake reconnaissance report, Bulletin of the NZSEE, Vol. 27, No. 4.
NZS 1170.5:2004. Structural design actions, Part 5: Earthquake actions - New Zealand, NZS 1170.5:2004.
Standards New Zealand, Wellington, New Zealand.
NZS 3404 Part 1:1997. Steel structures standard, incorporating Amendments 1 and 2, NZS 3404:1997.
Standards New Zealand, Wellington, NZ.
NZS 4203:1976. Code of practice for the general structural design and design loadings for buildings,
NZS 4203:1976. Standards New Zealand, Wellington, NZ.
Park, R., Billings, I.J., Clifton, G.C., Cousins, J., Filiatrault, A., Jennings, D.N., Jones, L.C.P., Perrin, N.D.,
Rooney, S.L., Sinclair, J., Spurr, D.D., Tanaka, H. and Walker, G. (1995). The Hyogo-ken Nanbu earthquake
(the Great Hanshin earthquake) of 17 January 1995, Report of the NZSEE Reconnaissance Team. Bulletin of
the New Zealand National Society for Earthquake Engineering Vol. 28(1): 1–98.
Roeder, C.W., Knechtel, B., Thomas, E., Vaneaton, A., Leon, R.T. and Preece, F.R. (1996). Seismic behaviour
of older steel structures, Journal of Structural Engineering, Vol. 122, No. 4, April 1996, 365-373.
Roeder, C.W., Leon, R.T. and Preece, F.R. (1994). Strength, stiffness and ductility of older steel structures
under seismic loading, University of Washington Department of Civil Engineering, Washington, USA, Report
No. SGEM 94–4.
Rondal, J. and Rasmussen, K. (2003). On the strength of cast iron columns, University of Sydney, Research
Report No. R829.
SB4.6 (2007). Improved assessment methods for static and fatigue resistance of old steel railway bridges,
Background document D4.6 to Guideline for load and resistance assessment of railway bridges. Prepared by
Sustainable Bridges - a project within EU FP6.
SCI (1995). Joints in steel construction – Moment connections, Steel Construction Institute, Ascot, U.K.,
Publication No. 207/95.
Storie, L.B., Pender, M.J., Clifton, G.C. and Wotherspoon, L. (2014). Soil-foundation-structure-interaction for
buildings on shallow foundations in the Christchurch earthquake, 10th US National Conference on Earthquake
Engineering, Anchorage, Alaska, NCEE.
Trahair, N.S., Bradford, M.A., Nethercot, D.A. and Gardner, L. (2008). The behaviour and design of steel
structures to EC3, 4th ed. Taylor & Francis. Abingdon, Oxon, UK.
Tremblay, R. and Robert, N. (2000). Seismic design of low- and medium-rise chevron braced steel frames,
Canadian Journal of Civil Engineering. 27: 1192–1206.
Wood, P.J. (1987). Cyclic testing of a concrete encased riveted beam column joint, Ministry of Works and
Development, Wellington, New Zealand, Report Number 5–82/2.
C6A.1 General
This section gives general guidance on the typical pre-1976 steel building systems used in
New Zealand.
Note:
This information is based on published material and details supplied by design engineers.
C6A.2.1 Beams
Beams were typically rolled steel joist (RSJ) sections. These are I-sections where the inside
face of the flanges is not parallel to the outside face, being at a slope of around 15%. This
makes the flanges thicker at the root radius than at the tips.
The flange slenderness ratios of RSJ sections are always compact when assessed to
NZS 3404:1997.
These beams were typically encased in concrete for fire resistance and appearance. This
concrete contained nominal reinforcement made of plain round bars or, sometimes, chicken
wire.
C6A.2.2 Columns
Columns formed from hot-rolled sections
These columns were either RSJs used as columns or box columns formed by connecting two
channels, toes out, with a plate to each flange. The columns were encased in lightly
reinforced concrete containing nominal reinforcement made of plain round bars.
Figure C6A.1: Riveted steel fabrication details, Government Life Insurance Building, 1937
(Wood, 1987)
Figure C6A.2: Riveted steel fabrication details, Government Life Insurance Building, 1937
(Wood, 1987)
The RSJ beam web was connected by a double clip angle connection to the column flanges,
also as shown in Figure C6A.2.
These joints generally involved the use of rivets up to 1950 and HSFG bolts after 1960, with
a changeover from rivets to bolts from 1950 to 1960.
Arc welding
Beam to column connections from about 1940 onwards were also arc welded.
The strength and ductility available from welded connections will need careful evaluation
and attention to load path. This topic is addressed in Section C6.6.1 and its importance is
illustrated in Figure C6A.3. This figure is taken from a building that collapsed in the Kobe
earthquake of 1995 (while this example is from Japan, the details are relevant to some early
New Zealand buildings and the concept is certainly relevant). It shows a failed beam to
column minor axis connection, forming part of a moment resisting frame in that direction.
The beam was welded to an endplate which was fillet welded to the column flange tips.
Unlike the connection detail shown in Figure C6A.2, there was no way to transfer the
concentrated axial force in the beam flanges induced by seismic moment reliably from the
beam into the column. As a result, the weld between endplate and column flange unzipped
under the earthquake action.
Figure C6A.3: Failed beam to column weak Figure C6A.4: Braced frame with light
axis welded connection from the 1995 tension bracing showing damage but no
Kobe earthquake collapse from the 1995 Kobe earthquake
Figures C6A.1 and C6A.2 show plated box columns connected by riveted angles.
Figure C6A.3 shows a bolted UC splice detail in the column, a forerunner of the bolted
column splice details of HERA Report R4-100 (Hyland, 1999). Such bolted splices generally
perform well.
Figure C6A.4 shows an X-braced CBF with relatively light bracing, while Figure C6A.5
shows a V-braced CBF. While both examples are from Kobe, Japan they have similar details
to early New Zealand buildings.
Figure C6A.5: V-braced CBF showing damage but no collapse from the 1995
Kobe earthquake
Note:
*A nominal yield strength of 210 MPa may be used for steel manufactured before 1948 in the UK.
Table C6B.3: Characteristic/nominal properties of high tensile structural steels from the UK
(Bussell, 1997 and Bates, 1991)
Period Plate thickness, Yield strength Tensile strength Ultimate strain
𝒕𝒕 (MPa) (MPa) (mm/mm)
(mm)
Table C6B.4: Characteristic/nominal properties of high tensile structural steels from the UK
manufactured to BS 4360:1968 (1968-86)
Grade Plate thickness, Yield strength Tensile strength Ultimate strain
𝒕𝒕 (MPa) (MPa) (mm/mm)
(mm)
C6B.2 Australia
The characteristic (lower)/nominal properties of steelwork provided in Australian standard
specifications before the introduction of joint AS/NZ standards in 1996 are given in
Tables C6B.5 and C6B.6.
A revision of AS 1204-1972 and AS 1405-1973. Grades 180 and 210 plates were replaced
by new grade 200 plates. Grades 300,400 and 500 plates were removed. The rest remained
the same as in AS 1204-1972.
BHP Australia replaced most of their Grade 250 sections with new Grade 300 sections in 1994,
while BHP New Zealand replaced their Grade 350 sections with new Grade 300 sections
Note:
*A nominal yield strength of 210 MPa may be used for steel manufactured before 1937 in Australia.
Table C6B.7: Characteristic/nominal strengths of mild structural steels to AS/NZS 3678 and
AS/NZS 3679
Period Grade Plate Yield Tensile Standard
thickness strength strength
(mm) (MPa) (MPa)
1996-now Same as 3678-1990, but a new grade 450 is added. AS/NZS 3678:1996
AS/NZS 3678:2011
450 𝑡𝑡 ≤ 20 450 520
2010-now Grade 250 and 400 sections removed. A new AS/NZS 3679:2010
S0 grade introduced. The rest remained the same (Plates)
as AS/NZS 3679:1996.
Structural steelwork imported from the USA before the 1960s is likely to have a lower yield
strength than that imported from the UK (refer to Table C6B.9). Geometric properties of
US sections can be obtained from publications such as that by Ferris (1954).
Table C6B.9: Characteristic/nominal strengths for steels manufactured in the USA for
buildings, based on Ferris (1954) and ASCE 41-13 (2014)
Period Yield strength Tensile strength
(MPa) (MPa)
Pre-1948 210 -
1948–Now 230 -